Relative Volume (RVOL)

A ratio comparing current volume to what is typical for that moment, so 'high volume' becomes a measured statement instead of a guess.

What it is

Relative Volume (RVOL) expresses a bar's volume as a multiple of a baseline: a reading of 2.0 means volume is running at roughly twice its typical level, and 0.5 means about half. The entire meaning of the number depends on what 'typical' means, and this is where many free implementations quietly go wrong. Intraday volume is not flat across the day — it clusters around session opens and closes, around session overlaps in forex, and around scheduled news. Calculating RVOL correctly for intraday charts means comparing the current bar with the average volume of the same time-of-day slot across prior sessions (time-bucketed RVOL), not with a simple rolling average of the last N bars (naive RVOL). The two versions can disagree substantially on the same bar. RVOL is descriptive: it summarizes how active the market has been relative to its own history and says nothing about future direction. This is educational material, not financial advice, and all trading carries risk of loss.

How it works

Naive RVOL divides the current bar's volume by a simple moving average of recent volume — for example, volume divided by its 20-bar SMA. That baseline mixes together bars from very different parts of the trading day. Because intraday volume typically follows a U-shape (heavy near the open and close, light midday, and heavy at session overlaps in FX), naive RVOL systematically reads 'high' near every open — even a perfectly ordinary open — and 'low' through every lunchtime lull. Time-bucketed RVOL removes that seasonality: for the 09:35 bar, average the volume of the 09:35 bar from each of the prior K sessions and divide today's 09:35 volume by that slot average. A reading of 2.0 then means twice the volume normal for this specific time of day, which is the question traders actually mean to ask. On 24-hour markets with regular bars, the same slot sits a fixed number of bars back each day, which makes the calculation deterministic. One further honesty note: spot forex has no centralized volume, so 'volume' on FX charts is tick volume from your specific feed — RVOL there measures activity at your broker, not global traded volume.

How traders read it

• 1.0 is the baseline: volume is running at its typical level for that slot; readings meaningfully above 1 describe unusually active conditions and readings below describe quiet ones.
• Many traders treat roughly 1.5-2.0+ as 'elevated participation' context around breakouts or news — it describes activity, not direction, and is not an instruction to act.
• Very low RVOL (for example below 0.5) describes thin conditions where ranges are often narrow and spreads can be relatively wider.
• Always check which baseline your indicator uses: a naive rolling average will read high at every session open by construction, while a time-bucketed baseline only flags opens that are unusual for an open.
• The current forming bar's RVOL climbs as the bar accumulates volume, so a mid-bar reading is not comparable to finished bars — read confirmed bars, or use an implementation that pro-rates elapsed time.
• RVOL confirms participation, never direction: a 3.0 reading occurs in breakouts, reversals, and failed moves alike.

Common settings

Time-bucketed versions typically average the same slot across 5-20 prior sessions (5-10 is common; more sessions smooth the baseline but adapt slower to regime changes). Naive versions usually divide by a 20-50 bar SMA of volume. On very low timeframes, a multi-day slot lookback can exceed the chart's available history, so shorter lookbacks are practical there. Equities and other session-gapped markets need session-aware bucketing rather than a fixed bar offset, and half-days, holidays, and DST shifts all distort the buckets.

Strengths

• Normalizes volume across time-of-day and across instruments, making 'busy' and 'quiet' comparable statements
• Simple, intuitive ratio: 1.0 is normal, 2.0 is double, 0.5 is half
• Useful context filter alongside price-based tools — participation behind a move is visible at a glance
• Fully deterministic and testable once the baseline is defined precisely
• Works on any market with volume data, on any intraday or daily timeframe

Pitfalls to watch

• Naive RVOL is systematically biased intraday: it reads high at every open and low at every lunchtime simply because the baseline ignores time-of-day seasonality.
• Forex volume is tick volume from your specific feed — broker-dependent and not a measure of global traded volume, so FX RVOL comparisons across feeds are unreliable.
• The forming bar's RVOL is incomplete by definition and rises through the bar; acting on partial-bar readings compares apples to oranges.
• Holidays, half-days, and daylight-saving shifts misalign the time buckets, producing distorted readings for days afterward.
• Settings sensitivity: lookback length and baseline choice change every reading — a threshold tuned on one instrument or session regime does not transfer automatically.
• High RVOL has no predictive power: it describes that participation is elevated, not whether the move will continue, and no indicator removes the risk of loss.

Pine v6 example

//@version=6
indicator("Time-of-Day RVOL vs Naive RVOL", overlay = false)

lookbackDays = input.int(5, "Slot Lookback (days)", minval = 1, maxval = 20)
naiveLen     = input.int(20, "Naive Average Length (bars)", minval = 1)

// Bars per day on regular intraday timeframes (24h markets like forex/crypto).
// Session-gapped markets (equities) need session-aware bucketing instead.
barsPerDay = math.max(1, math.round(86400.0 / timeframe.in_seconds()))

// Time-bucketed baseline: average volume at this SAME slot on prior days
float slotSum   = 0.0
int   slotCount = 0
for i = 1 to lookbackDays
    histVol = volume[i * barsPerDay]
    if not na(histVol)
        slotSum   += histVol
        slotCount += 1
slotAvg = slotCount > 0 ? slotSum / slotCount : na
rvolTod = not na(slotAvg) and slotAvg > 0 ? volume / slotAvg : na

// Naive baseline most free scripts use: a flat rolling average
naiveAvg  = ta.sma(volume, naiveLen)
rvolNaive = naiveAvg > 0 ? volume / naiveAvg : na

plot(rvolTod,   "RVOL (time-of-day)", color = color.teal)
plot(rvolNaive, "RVOL (naive)",       color = color.gray)
hline(1.0, "Baseline (typical volume)", color = color.gray)

Pro tip: Before trusting any RVOL number, ask one question: relative to what? Plot the time-bucketed and naive versions side by side for a week and watch how often they disagree at the open — that gap is measurement error, not market information. Read confirmed bars rather than the forming bar, and treat elevated RVOL as context about participation, never as a directional signal. This is educational content only, not financial advice; all trading carries risk of loss.

Educational only — not financial advice, not a recommendation to trade. No indicator is predictive; trading involves substantial risk of loss.